Sample input and Visualization of molecular orbital in molden

debu_molcas · 2016-11-13 11:37:55

Dear users,
I am new in molcas. I am reading manual but unfortunately I can't find the things. It would be very kind of you if you help me to learn the program.

1. Could you please provide a sample input for potential energy surface scan (e.g., Increase of bond distance with energy), TS and IRC calculation.

2. For the visualization of orbital I used the keyword ">export MOLCAS_MOLDEN=ON" and I can see the orbital properly for closed shell calculations. But when I have done the open shell calculations the test.guessorb.molden files prints only alpha orbital coefficients. Could you please tell me how to get the beta orbital coefficients?

I am using the following input

*
>export MOLCAS_MOLDEN=ON
 &GATEWAY
Coord
4
Coordinates in Angstrom units
C 0.000000 0.000000 0.118889
H 0.000000 1.008806 -0.237778
H -0.873651 -0.504403 -0.237778
H 0.873651 -0.504403 -0.237778
Basis=C.Pol.Sadlej.10s6p4d.5s3p2d.., H.6-31G..4s.2s.., 
Title=GateWay data By Gabedit
Group=Full
End of input

 &SEWARD
Title=Seward data By Gabedit
End of input

 &SCF
Title=SCF data By Gabedit
UHF
ZSpin=1
End of input

 &GRID_IT
ASCII
ALL
End of input
* remove the first star in the next file for get the cube files
*! cp *.cub*   $MOLCAS_SUBMIT_PWD/.

Your help will be highly appreciated.

Thanks & Regards

Debasish

Ignacio · 2016-11-14 11:39:37

A scan:

>>> FOREACH R in (0.9, 1.0, 1.1, 1.2)

&GATEWAY
  Coord
    2
    Angstrom
    H 0.0 0.0 0.0
    H 0.0 0.0 $R
  Group = NoSym
  Basis = STO-3G

&SEWARD

&SCF

>>> ENDDO

A TS:

&GATEWAY
  Coord
    3
    Angstrom
    H  0.000  1.070  0.000
    C  0.000  0.000  0.000
    N  1.150  0.000  0.000
  Group = NoSym
  Basis = STO-3G
  Constraints
    a = angle H1 C2 N3
   Values
    a = 60 degrees
  End of TSConstraints

>>> DOWHILE

&SEWARD

&SCF

&SLAPAF
  FindTS

>>> ENDDO

*** Have to run gateway again to fully disable the constraints

&GATEWAY
  Coord = $WorkDir/$Project.Opt.xyz
  Group = NoSym
  Basis = STO-3G

&SEWARD

&SCF

&MCKINLEY

An IRC:

&GATEWAY
  *** Geometry from TS calculation
  Coord
    3
    Angstrom
    H     0.09674913     1.08636568     0.00000000
    C    -0.07663786    -0.10308840     0.00000000
    N     1.12988873     0.08672272     0.00000000
  Group = NoSym
  Basis = STO-3G

*** Need some additional iterations (default is 50)
>>> EXPORT MOLCAS_MAXITER = 100
>>> DOWHILE

&SEWARD

&SCF

&SLAPAF
  IRC
  *** Reaction vector from imaginary frequency mode
  Reaction vector
    0.99047  0.07207 0.00000
   -0.04428 -0.08107 0.00000
   -0.03334  0.06429 0.00000

>>> ENDDO

As for the orbitals, why do you open the *.guessorb.molden file? That's, as its name says, guess orbitals, which have not been optimized at all (and at that point of the calculation the program didn't even know it was going to be UHF. Open the *.scf.molden file and all should be good.

debu_molcas · 2016-11-16 15:10:05

Thank you very much for your reply. I have few more queries.

1. Is it possible to do relaxed PES scan in molcas. if possible please provide me a sample input.

2. I have run a frequency calculation and in the output it printed the frequency but not the displacement vector. although the calculation TS.inp which you have provided prints the vectors. I have compared the inputs but did not find any difference. Could you please see the following input and let me know what is the problem?

* 
&GATEWAY 
Basis set
O.6-31G*....
O1 0.000000 0.000000 0.070437 /Angstrom
End of basis
Basis set
H.6-31G*....
H1 0.832080 0.000000 -0.558946 /Angstrom
End of basis
Title=GateWay data By Gabedit
Symmetry
XY Y 
End of input
>>>  Export  MOLCAS_MAXITER=100 
>>>  Do  while 

  &SEWARD 
  &SCF;  
  KSDFT=B3LYP
  &ALASKA &END
  End of input
  &SLAPAF 

>>>  EndDo 

*&Grid_It
&MCKINLEY

one more problem I have found if I enable the Grid_IT module. It is not produced the grid file and terminates with the following error

Irreps  :     1    2    3    4
Basis   :    10    5    1    3
Grid Net:    30   24   24

 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    Location: rdVec_                                                     ###
 ###    File: INPORB                                                         ###
 ###                                                                         ###
 ###                                                                         ###
 ###    NSYM does not match                                                  ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################
[ process      0]: xquit (rc =    128): _INTERNAL_ERROR_
--- Stop Module:  grid_it at Wed Nov 16 16:08:20 2016 /rc= _INTERNAL_ERROR_ ---
*** files:   H2O.grid
    saved to directory /home/sason/debasish/MOLCAS/TEST-D/freq

Non-zero return code - check program input/output
--- Stop Module:  auto at Wed Nov 16 16:08:20 2016 /rc= _INTERNAL_ERROR_ ---

Could you please let me know what is the problem?

Ignacio · 2016-11-16 17:51:08

debu_molcas wrote:

1. Is it possible to do relaxed PES scan in molcas. if possible please provide me a sample input.

Combine a FOREACH loop with a constrained optimization:

>>> FOREACH ANGLE in (90, 100, 110)

&GATEWAY
  Coord
    3
    Angstrom
    O 0.0 0.0 0.0
    H 1.0 0.0 0.0
    H 0.0 1.0 0.0
  Group = NoSym
  Basis = STO-3G
  Constraints
    a = Angle H2 O1 H3
  Values
    a = $ANGLE degrees
  End of Constraints
  
>>> DOWHILE

&SEWARD

&SCF

&SLAPAF

>>> ENDDO

>>> ENDDO

2. I have run a frequency calculation and in the output it printed the frequency but not the displacement vector. although the calculation TS.inp which you have provided prints the vectors. I have compared the inputs but did not find any difference. Could you please see the following input and let me know what is the problem?

There are several differences. Apart from the molecule and the quantum method, you run with symmetry (which results in numerical frequencies), and it is not a transition state optimization. But I get the frequencies and modes:

 Harmonic frequencies in cm-1

 IR Intensities in km/mol

                        1         2

     Frequency:      1713.60   3726.08

     Intensity:    0.757E+02 0.170E+01

     O1         z   -0.06785   0.04892
     H1         x    0.41194   0.57133
     H1         z    0.53842  -0.38821

However, these are only the fully symmetric modes, there's no asymmetric stretch because you run with symmetry.

one more problem I have found if I enable the Grid_IT module. It is not produced the grid file and terminates with the following error

Your input, with grid_it enabled, runs fine here. Which molcas version are you using?

debu_molcas · 2016-11-17 01:25:33

Thanks a lot for the reply. I got the logic of relaxed PES scan input.

About Grid_IT module: The new calculation I have done for a closed shell system successfully prints the grid file. But the input I have send in my last mail is terminates with error. I am using Molcas version 8.1 patch level 160404-1800.

About Frequency calculation: Numerical or Analytical which one is default for frequency calculation? If I want to change the method what will be the keyword for it?

If I use symmetry in calculation it will only calculate the symmetric modes. So, to get all the modes we need to switch of the symmetry. But if the molecule has some symmetry we should always use it for the calculation. Is it correct?

Now I have few more queries. Actually I would like to do a reaction modelling using CASSCF/CASPT2 method. I have optimized the molecule (closed Shell) in DFT/B3LYP method. Can I use this as a initial guess of orbital for CAS calculation? Could you please suggest me HF or DFT which one is better for initial guess and geometry for CAS calculation?

I want to use 8 active electrons and 8 orbitals (4 occupied & 4 unoccupied) in active space. But the active molecular orbitals in the aforementioned calculations are not in exactly HOMO-3 to LUMO+3. So, I need to alter the orbitals. Could you please help me little more here to prepare a input to start a CAS optimization after alter the MOs using the initial guess?

One more question: I have seen in manual that molcas has an option for calculation in presence of external electric field. Is it possible to use for all the method e.g., CASSCF/CASPT2 ? What is the keyword to use it?

If I don't need Isotopic shifts data is there any way to stop this during frequency calculation?

I look forward for your kind response and I am sorry for this large mail.

Thanks & Regards

Debasish

Ignacio · 2016-11-17 09:52:18

Please do not add more and more questions to the same topic. If you have new questions, create a new topic.

debu_molcas wrote:

About Grid_IT module: The new calculation I have done for a closed shell system successfully prints the grid file. But the input I have send in my last mail is terminates with error. I am using Molcas version 8.1 patch level 160404-1800.

That's a development version, it means you are a developer or are collaborating with one. I've run it with 161111-0800 and it works fine. Anyway, the error says the symmetry has changed, maybe you have a dirty WorkDir.

About Frequency calculation: Numerical or Analytical which one is default for frequency calculation? If I want to change the method what will be the keyword for it?

Molcas will compute analytical frequencies if available, otherwise it will fall back to numerical. You can force numerical gradients and frequencies with the keyword "Numerical" in SEWARD. You cannot force analytical frequencies if they are not available.

If I use symmetry in calculation it will only calculate the symmetric modes. So, to get all the modes we need to switch of the symmetry. But if the molecule has some symmetry we should always use it for the calculation. Is it correct?

No, you can perfectly calculate a symmetrical molecule without using symmetry. You just have to include the complete structure in the input (not just the symmetry-unique atoms).

Now I have few more queries. Actually I would like to do a reaction modelling using CASSCF/CASPT2 method. I have optimized the molecule (closed Shell) in DFT/B3LYP method. Can I use this as a initial guess of orbital for CAS calculation? Could you please suggest me HF or DFT which one is better for initial guess and geometry for CAS calculation?

The geometry will typically be better with DFT (depending on the functional). The orbitals will probably be of similar quality, you can use any of them as initial guess for CASSCF.

I want to use 8 active electrons and 8 orbitals (4 occupied & 4 unoccupied) in active space. But the active molecular orbitals in the aforementioned calculations are not in exactly HOMO-3 to LUMO+3. So, I need to alter the orbitals. Could you please help me little more here to prepare a input to start a CAS optimization after alter the MOs using the initial guess?

You can use the keyword "Alter" in RASSCF to swap a number of orbital pairs. You can also modify the end of the .ScfOrb file (the #INDEX block) and write a "2" instead of "i" or "s" in the orbitals you want to use, and then use this file in RASSCF through the "File" keyword.

One more question: I have seen in manual that molcas has an option for calculation in presence of external electric field. Is it possible to use for all the method e.g., CASSCF/CASPT2 ? What is the keyword to use it?

Yes, as far as I know. Check the "XField" keyword in GATEWAY and the FFPT program.

If I don't need Isotopic shifts data is there any way to stop this during frequency calculation?

I don't think so.

debu_molcas · 2016-11-17 12:43:10

Thank you very much for nice explanations. I will post my other queries in respective section.

Now I have one questions that may be related to this section. I have tried to optimize a anthracene molecule for test in RHF/STO-3G level of theory.

Here is the input

> Export  MOLCAS_MAXITER=500
> Export  MOLCAS_PRINT=3
>>> Do while

 &GATEWAY
Coord
24
Coordinates in Angstrom units
 C                 -5.07670324   -2.30811820   -0.01291600
 C                 -3.71208024   -2.30811820   -0.01291600
 C                 -2.97918924   -1.07689720   -0.01291600
 C                 -3.70318624    0.15484980   -0.01843400
 C                 -5.13542624    0.11335880   -0.01973400
 C                 -5.79920924   -1.07893420   -0.01636300
 C                 -1.58028824   -1.05400820   -0.00802000
 C                 -3.00280924    1.36601680   -0.02015500
 C                 -1.60386824    1.38890080   -0.01453000
 C                 -0.87989924    0.15719780   -0.00711000
 C                  0.55228276    0.19882180    0.00000000
 H                  1.09815276   -0.75670520    0.00660100
 C                  1.21601976    1.39117680    0.00104200
 C                  0.49353676    2.62029180   -0.00713000
 C                 -0.87110624    2.62014680   -0.01449700
 H                 -1.02258524   -2.00324520   -0.00370100
 H                 -5.64257524   -3.25139620   -0.01051200
 H                 -3.14275424   -3.24985220   -0.00981000
 H                 -5.68123924    1.06890480   -0.02433800
 H                 -6.89860724   -1.11509920   -0.01779400
 H                 -3.56094824    2.31499780   -0.02442300
 H                  2.31542576    1.42708080    0.00720200
 H                  1.05961676    3.56346880   -0.00603500
 H                 -1.44064924    3.56176380   -0.02076900
Basis=C.STO-3G.., H.STO-3G.., 
Title=GateWay data By Gabedit
Group=NoSym
End of input

 &SEWARD
Title=Seward data
End of input

 &SCF
Title=SCF data
End of input

 &ALASKA 
 &SLAPAF

>>> EndDo

 &MCKINLEY

 &GRID_IT
ASCII
ALL
End of input

Unfortunately, even after 100 steps it is not optimizing. It is a very stable organic molecule and I hope the initial geometry is also not bad then why the geometry is not converging?

Could you please see the files and let me know what is the problem?

Thanks & Regards
Debasish

Ignacio · 2016-11-17 16:49:00

You have &GATEWAY inside the loop, that effectively starts from scratch every iteration, so you are doing 500 times the same calculation. You should start every iteration only with &SEWARD (which picks the geometry from the initial &GATEWAY or the previous &SLAPAF):

&GATEWAY
...
>>>Do While
&SEWARD
&SCF
...
&SLAPAF
>>>End Do

valera · 2016-11-30 17:15:45

Not exactly the same. SCF will reuse the density from the previous calculation smile

Molcas Forum

Announcement

#1 2016-11-13 11:37:55

Sample input and Visualization of molecular orbital in molden

#2 2016-11-14 11:39:37

Re: Sample input and Visualization of molecular orbital in molden

#3 2016-11-16 15:10:05

Re: Sample input and Visualization of molecular orbital in molden

#4 2016-11-16 17:51:08

Re: Sample input and Visualization of molecular orbital in molden

#5 2016-11-17 01:25:33

Re: Sample input and Visualization of molecular orbital in molden

#6 2016-11-17 09:52:18

Re: Sample input and Visualization of molecular orbital in molden

#7 2016-11-17 12:43:10

Re: Sample input and Visualization of molecular orbital in molden

#8 2016-11-17 16:49:00

Re: Sample input and Visualization of molecular orbital in molden

#9 2016-11-30 17:15:45

Re: Sample input and Visualization of molecular orbital in molden

Board footer